damage of concrete with epoxy coated reinforcement …

DAMAGE OF CONCRETE WITH EPOXY COATED REINFORCEMENT DUE TO CORROSION, FREEZE-THAW AND FATIGUE WITH

APPLICATION TO BRIDGE DECKS

By

Matthew Garratt

A THESIS

Submitted to Michigan State University

in partial fulfillment of the requirements for the degree of

MASTER OF SCIENCE

Civil Engineering

2011

ABSTRACT

DAMAGE OF CONCRETE WITH EPOXY COATED REINFORCEMENT DUE TO CORROSION, FREEZE-THAW AND FATIGUE WITH

APPLICATION TO BRIDGE DECKS

By

Matthew Garratt

The deterioration of reinforced concrete bridge decks due to corrosion of the steel bars is a

common problem. Epoxy coated reinforcement (ECR) is known to perform better than plain

(black) steel, but the appropriate maintenance schedule for ECR bridge decks is not well

established. The goal of this project was to compare damage of concrete reinforced with ECR

and black steel due to corrosion, freeze-thaw, and fatigue loading for application to bridge decks.

Concrete specimens with ECR and black steel reinforcement were artificially aged using

freeze-thaw cycling and accelerated corrosion to simulate the varying ages of bridge decks.

These specimens were repaired and then simultaneously subjected to fatigue loading, freeze-

thaw cycles, and continued accelerated corrosion to simulate the loading on bridge decks.

Two methods were used to compare the deterioration of specimens with black steel and ECR.

The first method was based on corrosion data. It was found that the corrosion rate was about 2.5

times lower for ECR than for black steel, implying that the effects of corrosion-induced damage

appear more slowly in ECR. The second method utilized X-ray tomography to obtain images of

internal cracking in the specimens. Image-processing software was used to compute the volume

of cracks inside the specimens. It took about 4 times longer for the ECR specimens to have the

same volume of cracks as black steel specimens. The damage was insensitive to the type of

repair performed.

iii

ACKNOWLEDGMENTS

First and foremost, I give thanks to the Lord Jesus Christ. This thesis would not have been

possible without the strength and knowledge given to me by Him.

I also want to thank my advisor, Dr. Ron Harichandran for giving me the opportunity work

on this research. I cannot thank him enough for his financial support for graduate school, advice

on this project, and guidance for this thesis. I would also like to thank Dr. Rigoberto Burgueño

for his advice and suggestions on this research. His passion and professionalism in the classroom

as a teacher encouraged me to want to pursue an advanced degree and to become involved in

research.

The Michigan Department of Transportation provided funding for this research. Their

support is gratefully acknowledged. My fellow colleague Dr. Gang Zhang and I collaborated in

completing this project. I thank him for his advice and help in conducting this research. Finally, I

want to thank Siavosh Ravanbakhsh for his help in the lab conducting the experiments

throughout this project. His expertise was invaluable to this research and his humor made those

tough times in the lab tolerable.

iv

TABLE OF CONTENTS

LIST OF TABLES ...........................................................................................................................v

LIST OF FIGURES ...................................................................................................................... vii

CHAPTER 1: INTRODUCTION ....................................................................................................1

1.1 Corrosion Mechanism of Reinforced Concrete Bridge Decks ..........................................2

1.2 Corrosion Control Methods ...............................................................................................5

1.3 Corrosion Performance of Epoxy Coated Reinforcement.................................................7

1.3.1 Corrosion Performance of ECR in Laboratory Tests ................................................ 7

1.3.2 Corrosion Performance of ECR in Concrete Bridge Decks ...................................... 9

1.3.3 Summary of the Section ........................................................................................ ..13

1.4 Repair Options for Reinforced Concrete Bridge Decks ..................................................14

1.5 Summary .........................................................................................................................16

CHAPTER 2: SPECIMEN PREPARATION ................................................................................18

2.1 Specimen Design and Casting Preparation .....................................................................18

2.2 Accelerated Freeze-Thaw Test ........................................................................................25

2.3 Accelerated Corrosion .....................................................................................................27

2.4 Specimen Repair .............................................................................................................31

CHAPTER 3: EXPERIMENTAL DESIGN ..................................................................................37

3.1 Fatigue Testing ................................................................................................................37

3.2 Static Testing ...................................................................................................................41

CHAPTER 4: EXPERIMENTAL TEST RESULTS .....................................................................47

4.1 Static Test Results ...........................................................................................................47

4.2 Corrosion during Fatigue Loading ..................................................................................53

CHAPTER 5: X-RAY COMPUTED TOMOGRAPHY ...............................................................54

5.1 X-Ray Scanning ..............................................................................................................54

5.2 Image Processing.............................................................................................................56

CHAPTER 6: INTERPRETATION OF RESULTS AND CONCLUSIONS ...............................61

6.1 Corrosion Damage...........................................................................................................61

6.2 Image-Based Damage Results .........................................................................................66

6.3 Summary and Conclusions ..............................................................................................68

6.4 Final Remarks .................................................................................................................70

APPENDIX A: STIFFNESS DATA .............................................................................................73

APPENDIX B: X-RAY IMAGES AND PROCESSED CRACK IMAGES ................................77

APPENDIX C: CORROSION CHARGE DATA .........................................................................85

REFERENCES ..............................................................................................................................94

v

LIST OF TABLES

Table 2.1: Test Matrix...................................................................................................................20

Table 2.2: Corrosion Charge for Different Damage Levels .........................................................29

Table 3.1: List of Groups of Specimens .......................................................................................39

Table 4.1: Survival Under Fatigue Loading .................................................................................52

Table 5.1: Image-Based Damage ..................................................................................................60

Table 6.1: Estimates of Time in the Field for Black Steel ............................................................62

Table 6.2: Time Scale Factors from Image-Based Damage .........................................................67

Table A.1: Black Level 0 ..............................................................................................................73

Table A.2: Epoxy Level 0 Unrepaired ..........................................................................................73

Table A.3: Black Level 1 Unrepaired ...........................................................................................73


Table A.5: Black Level 1 Repaired...............................................................................................74

Table A.6: Epoxy Level 1 Repaired .............................................................................................74

Table A.7: Black Level 2 Unrepaired ...........................................................................................74


Table A.9: Black Level 2 Repaired...............................................................................................74

Table A.10: Epoxy Level 2 Repaired ...........................................................................................74

Table A.11: Black Level 3 Unrepaired .........................................................................................75

Table A.12: Epoxy Level 3 Unrepaired ........................................................................................75

Table A.13: Black Level 3 Repaired.............................................................................................75


Table A.15: Black Level 4 Unrepaired .........................................................................................75

Table A.16: Epoxy Level 4 Unrepaired ........................................................................................75

Table A.17: Black Level 4 Repaired.............................................................................................76


Table C.1: B-0 Corrosion Data ......................................................................................................85

vi

Table C.2: E-0 Corrosion Data ......................................................................................................85

Table C.3: B-1-N Corrosion Data ..................................................................................................85

Table C.4: E-1-N Corrosion Data ..................................................................................................86

Table C.5: B-1-R Corrosion Data ..................................................................................................86

Table C.6: E-1-R Corrosion Data ..................................................................................................86

Table C.7: B-2-N Corrosion Data ..................................................................................................86

Table C.8: E-2-N Corrosion Data ..................................................................................................87

Table C.9: B-2-R Corrosion Data ..................................................................................................87

Table C.10: E-2-R Corrosion Data ................................................................................................88

Table C.11: B-3-N Corrosion Data ................................................................................................88

Table C.12: E-3-N Corrosion Data ................................................................................................89

Table C.13: B-3-R Corrosion Data ................................................................................................89


Table C.15: B-4-N Corrosion Data ................................................................................................90

Table C.16: E-4-N Corrosion Data ................................................................................................91

Table C.17: B-4-R Corrosion Data ................................................................................................91


vii

LIST OF FIGURES

Figure 2.1: Specimen Configuration .............................................................................................19

Figure 2.2: ECR with Damage in Coating ....................................................................................21

Figure 2.3: Dimensions and Reinforcement Layouts of Pilot Beams ...........................................22

Figure 2.4: Load-Displacement Response of Pilot Test Beams ...................................................23

Figure 2.5: Failure Patterns of Pilot Test Beams ..........................................................................24

Figure 2.6: Molds before Casting .................................................................................................25

Figure 2.7: Specimens after Freeze-Thaw Test ............................................................................26

Figure 2.8: Accelerated Corrosion Test Setup ..............................................................................28

Figure 2.9: Corrosion Curves for BSR and ECR ..........................................................................29

Figure 2.10: Specimen after Level 1 Corrosion ............................................................................30




Figure 2.14: Specimen Repaired with Epoxy Overlay .................................................................32

Figure 2.15: Oversized Shallow Concrete Overlay ......................................................................33

Figure 2.16: Specimen with Shallow Concrete Overlay ...............................................................34

Figure 2.17: Specimen with Waterproofing and HMA Overlay ...................................................35

Figure 2.18: Oversized Deep Concrete Overlay ...........................................................................35

Figure 2.19: Specimen with Deep Concrete Overlay ...................................................................36

Figure 3.1: Environmental Chamber.............................................................................................38

Figure 3.2: Fatigue Test Setup ......................................................................................................39

Figure 3.3: Corrosion Setup ..........................................................................................................41

Figure 3.4: Static Test Setup .........................................................................................................43

Figure 3.5: Smooth Surface for LVDT .........................................................................................43

Figure 3.6: Data Obtained from Multiple Static Tests ..................................................................45

viii

Figure 3.7: Schematic of Static Test .............................................................................................45

Figure 3.8: Typical Load-Displacement Graph ............................................................................46

Figure 4.1: Normalized Stiffness Values (Level 0 Unrepaired) ...................................................48


Figure 4.3: Normalized Stiffness Values (Level 1 Repaired) .......................................................49






Figure 4.9: Charge vs. Fatigue Cycles ..........................................................................................53

Figure 5.1: X-Ray Scanning Setup ...............................................................................................55

Figure 5.2: Example of X-Ray Scan .............................................................................................55

Figure 5.3: Photograph of Scanned Specimen ..............................................................................56

Figure 5.4: Crack Cropped from Image ........................................................................................57

Figure 5.5: Crack with Inverted Colors ........................................................................................57

Figure 5.6: Crack with Enhanced Brightness and Contrast ..........................................................58

Figure 5.7: Crack after Applying Bandpass Filter ........................................................................58

Figure 5.8: Crack after Applying Threshold .................................................................................59

Figure 5.9: 3D Rendering of Crack...............................................................................................60

Figure 6.1: Charge vs. Time for Phase 1 and Phase 2 ..................................................................63

Figure 6.2: Residual and Weighted Residual Plots for Black Steel ..............................................64

Figure 6.3: Residual and Weighted Residual Plots for ECR ........................................................64

Figure 6.4: Early Stage of Accelerated Corrosion for ECR..........................................................65

Figure 6.5: Image-Based Damage vs. Time ..................................................................................67

Figure 6.6: Residual Plot with All Points Included ......................................................................68

Figure B.1: B-0 X-Ray Image and Crack after Applying Threshold ............................................77

ix

Figure B.2: E-0 X-Ray Image with No Crack ...............................................................................77

Figure B.3: B-1-N X-Ray Image and Crack after Applying Threshold ........................................78

Figure B.4: E-1-N X-Ray Image with No Crack ...........................................................................78

Figure B.5: B-1-R X-Ray Image and Crack after Applying Threshold .........................................79

Figure B.6: E-1-R X-Ray Image with No Crack ...........................................................................79

Figure B.7: E-2-N X-Ray Image and Crack with No Crack ..........................................................80

Figure B.8: B-2-R X-Ray Image and Crack after Applying Threshold .........................................80

Figure B.9: E-2-R X-Ray Image and Crack after Applying Threshold .........................................81

Figure B.10: B-3-N X-Ray Image and Crack after Applying Threshold ......................................81

Figure B.11: E-3-N X-Ray Image and Crack after Applying Threshold .......................................82

Figure B.12: B-3-R X-Ray Image and Crack after Applying Threshold .......................................82

Figure B.13: E-3-R X-Ray Image and Crack after Applying Threshold .......................................83

Figure B.14: B-4-R X-Ray Image and Crack after Applying Threshold .......................................83

Figure B.15: E-4-R X-Ray Image and Crack after Applying Threshold .......................................84

1

CHAPTER 1

INTRODUCTION

The highway system in the United States is very complex and plays an important role in its

economics. Bridges are built to span a valley, road, body of water, or other physical obstacles

and account for a substantial component in highway infrastructure. In bridge construction,

concrete has the advantages of being a durable, versatile, and economical material and is of great

interest for transportation agencies and bridge owners. Although the deterioration of structures

over time is normal, concrete was thought to be a relatively low maintenance material before the

first use of de-icing salts in the 1950s. However, when de-icing salts began to be applied in

northern locations in the winter, reinforcement corrosion in concrete bridges increased rapidly

[1]. Bridges in coastal locations have also been severely corroded, due to seawater or spray

exposure. A recent study [2] showed that 14 percent of the nation’s concrete bridges were rated

structurally deficient, and the primary cause of the deficiency was corrosion of the reinforcing

steel. The study determined that the annual cost of corrosion to all bridges (including steel

bridges) is $8.29 billion, including indirect costs incurred by bridge closures.

Due to the severity of the problem, researchers have attempted to develop corrosion

protection methods. Early research by the National Bureau of Science (now National Institute of

Standards and Technology) and the Federal Highway Administration (FHWA) indicated that

epoxy coated reinforcement (ECR) performed well in concrete contaminated by salts [3] and

provided protection such that the premature deterioration of concrete due to expansive corrosion

is minimized. However, field investigations found that even though ECR generally performs

better than traditional bare steel reinforcement (BSR), some bridges constructed using ECR are

2

now starting to show indications of surface damage and repairs are needed to prevent the

progress of the damage. It is not clear whether the repair options used for bridges with black (or

bare) steel reinforcement should be applied to bridges with ECR. It is therefore necessary to

investigate the frequency and type of repair that should be performed for ECR bridge decks.

This section first introduces the corrosion mechanism of reinforcement in bridge decks and

commonly used corrosion control methods. Research on the performance and applications of

ECR, including laboratory tests and field investigations, is then summarized. Research on the

repair of ECR bridge decks is also included.

1.1 Corrosion Mechanism of Reinforced Concrete Bridge Decks

The corrosion of steel reinforcement is an electro-chemical process in which the steel reacts

with the surrounding environment and the metal is converted into a metal oxide compound. For

this process to occur there has to be a current flow, which results from a potential difference

between two nodes, typically on the same reinforcing bar. The water inside the concrete is

usually alkaline and this protects the steel because a protective oxide (passive) film forms under

this condition. The exact product may vary with the pH value and can be expressed in a general

form as [4]:

m · Fe(OH)2 + n · Fe(OH)3 + p · H2O (1.1)

This film effectively protects the steel so that the corrosion rate is negligible, allowing

decades of relatively low maintenance [5].

However, when there is a change in the environment, for example, a decrease in the pH value,

corrosion of the steel will start. Two mechanisms commonly induce corrosion in the steel

3

reinforcement. One mechanism is carbonation, where carbon-dioxide dissolved in the water

reacts with calcium ions in the cement and produces calcium carbonate:

CaOH2 + CO2→ CaCO3 + H2O (1.2)

The loss of OH- reduces the pH of the water in the concrete. When the pH reduces to a

certain level, the steel is no longer passive and will start reacting with the oxygen dissolved in

the water. The associated reactions are:

At anode:

3Fe+8OH-→Fe3O4+4H2O+8e- (1.3)

At cathode:

O2+2H2O+4e-→4OH- (1.4)

The rate of this type of corrosion depends on factors including the availability of water and

oxygen and the surface area exposed. The amount of oxygen in the concrete depends on the

diffusion rate and is affected by the water saturation in the concrete. When fully saturated, the

diffusion rate is low. When the concrete is dry, the oxygen can move freely in the pores, but the

reactions cannot take place due to the lack of water. Wet and dry cycles accelerate the corrosion

process as it allows oxygen to diffuse and the water to act as an electrolyte.

A second mechanism of corrosion in concrete reinforcement is chloride-induced corrosion in

which chloride ions from de-icing salts or marine environments can penetrate the concrete

4

through cracks and destroy the passive film on the steel. The product of the chemical reaction is

a soluble iron-chloride complex [6]. The corrosion reaction at the anode is:

Fe→Fe2++2e- (1.5)

The reaction at the cathode is the same as that in carbonation-induced corrosion. If the supply

of water and oxygen is sufficient, the iron-chloride will react with oxygen and become Fe2O3 or

Fe3O4 and release the chloride to continue the corrosion process. A low concentration of 0.6

kilograms per cubic meter (concrete weight basis) can compromise steel passivity [5].

Considerable efforts have focused on identifying a chloride threshold for initializing corrosion.

However, a unique value for this parameter has remained unavailable because many factors such

as cement type, concrete mix design, environmental factors, and reinforcement composition can

influence this value.

Figure 1.1 shows corroded reinforcing bars in concrete. The volume of corrosion products for

both types of corrosion is approximately 3-6 times the volume of the original steel. Tensile

stresses will develop due to the corrosion-induced expansion and results in cracks, delaminations,

and spalls in concrete due to its low tensile strength. This further accelerates the corrosion

process by providing more pathways for water, oxygen, and chloride ions.

5

Figure 1.1: Steel Corrosion in Concrete. For interpretation of the references to color in this and all other figures, the reader is referred to the electronic

version of this thesis.

1.2 Corrosion Control Methods

To prevent or reduce the damage caused by corrosion, different corrosion control methods

have been investigated and developed. This section briefly describes the commonly used

methods.

One method is to use stainless steel or stainless steel-clad reinforcement. Stainless steel

differs from regular black steel. A passive film of chromium oxide forms on stainless steel which

prevents further surface corrosion and blocks the spread of the corrosion into the internal

structure [7]. The corrosive threshold of stainless steel in concrete is about 10 times higher than

that for traditional black steel [8]. Field investigations have found that bridge decks using

stainless-reinforcement had no corrosion induced cracks after 9 years of service. This makes

stainless steel an attractive material to reduce corrosion in highway and bridge infrastructure.

However, the cost of stainless steel is a major concern for owners. The use of stainless steel in a

6

bridge deck increases the initial cost by 5.5 to 15.6 percent [9], and is therefore typically

considered to protect only critical and hard-to-repair structures.

Another method to control or reduce corrosion of reinforcing bars in concrete is cathodic

protection. In cathodic protection, an external current is supplied to the protected metal by an

auxiliary anode. As a result, corrosion is reduced or stopped. Two methods are used to supply

external current. In one method, the protected metal is connected to a more active sacrificial

metal, for example, zinc. In the second method, an external current source is applied [10]. This

method is effective in reducing the corrosion of steel in concrete, but it has several drawbacks. If

an external power source is used, the power source needs to be charged and replaced regularly.

The wiring and connections also induce additional cost. If an active metal is used, the cost of

obtaining and replacing the sacrificed metal can also be high.

Epoxy-coated reinforcing bars is one of the most widely used anti-corrosion methods. The

epoxy coating reduces corrosion by [11]:

• Resistance inhibition: providing electrical resistance to limit current transfer between

anodic and cathodic sites.

• Oxygen deprivation: excluding oxygen and thereby impeding the cathodic reaction.

• Inhibition or aestivation: introducing material into the interfacial environment to

stimulate the development of passive or inhibitive surface films.

Even though the cost of ECR is higher than that of regular black steel, the benefit of ECR is

obvious and it is cost-effective in the long run. Ideally, bridge decks with ECR should not suffer

from corrosion-induced damage. However, in reality, corrosion of ECR does occur. For example,

when the adhesion between the epoxy coating and the reinforcement is lost, the surface of the

7

steel will be exposed to the environment and corrosion is likely to occur. Water absorption

properties of the coating can influence the de-bonding of the coating. Another cause for

corrosion of ECR is damage to the coating during manufacture, fabrication, transportation, or

construction. Small regions of steel exposed to the surrounding environment can corrode. Even

though these problems can be reduced by proper design of the coating and proper handling

during the transportation and construction process, they cannot be completely eliminated and the

signs of ECR corrosion have been observed in both field investigations and laboratory

experiments.

1.3 Corrosion Performance of Epoxy Coated Reinforcement

Extensive research has been conducted in the past decade in order to understand the

performance of ECR. This section summarizes the results of previous research and is divided

into two sub-sections: performance of ECR in laboratory tests and field performance of ECR in

concrete structures.

1.3.1 Corrosion Performance of ECR in Laboratory Tests

To compare the corrosion performance of different types of bars, the FHWA [12] evaluated

the corrosion of organic, ceramic, inorganic, and metallic clad and solid metallic bars. Corrosion

tests of the bars themselves, as well as tests of the bars in concrete, were performed. The results

showed that black bars exhibited very poor corrosion performance. Galvanized bars had a better

performance than black steel bars when such bars were used for the entire structure. The

corrosion rate of copper clad bars was 95 percent lower than that of black steel bars. The

corrosion rate of Type 304 stainless steel bars varied with the cathode: if the cathode was

stainless the corrosion rate was 99.8 percent lower than that of the black steel bars, while the rate

became 90-95 percent if black steel

no sign of corrosion and had the best performance.

concrete and the amount of damage to the b

corrosion rate of ECR was 100 times less than that of black steel

layers were ECR. The type of coating also played a role but the difference reduced if

were epoxy coated. According to the results of the report

system but several requirements are needed to maximize

used for the entire structure. Second,

damage to the coating needs to be minimized.

Researchers at Wiss, Janney, Elstner Associates, Inc

corrosion test to compare the performance of ECR, iron chromium alloy, galvanized bars

stainless steel bars. They found that ECR performed very well and corroded at drilled holes

shown in Figure 1.2. ECR removed from a 15

Corrosion was minor and only occurred

also observed that bars with poor coat bond still performed

bonding may not affect the corrosion performance significantly.

Figure

8

95 percent if black steel bars were used as cathodes. Type 316 stainless

the best performance. For ECR, the presence of cracks in the

concrete and the amount of damage to the bars played a significant role in the performance.

corrosion rate of ECR was 100 times less than that of black steel bars when both top and bottom

. The type of coating also played a role but the difference reduced if

According to the results of the report, ECR is a good corrosion protection

system but several requirements are needed to maximize its performance. First, ECR needs to be

r the entire structure. Second, the cracks in the concrete need to be repaired

needs to be minimized.

Wiss, Janney, Elstner Associates, Inc. (WJE) [13] conducted an accelerated

corrosion test to compare the performance of ECR, iron chromium alloy, galvanized bars

found that ECR performed very well and corroded at drilled holes

. ECR removed from a 15-year old deck performed well during the test.

occurred at drilled holes or at existing defects in the coating.

also observed that bars with poor coat bond still performed well, indicating that the effect of coat

not affect the corrosion performance significantly.

Figure 1.2: ECR after Corrosion

. Type 316 stainless steel showed

For ECR, the presence of cracks in the

ars played a significant role in the performance. The

when both top and bottom

. The type of coating also played a role but the difference reduced if all bars

, ECR is a good corrosion protection

performance. First, ECR needs to be

epaired. Third, the

conducted an accelerated

corrosion test to compare the performance of ECR, iron chromium alloy, galvanized bars, and

found that ECR performed very well and corroded at drilled holes as

year old deck performed well during the test.

s in the coating. They

, indicating that the effect of coat

9

NFESC (Naval Facilities Engineering Service Center) performed a test to compare the

performance of ECR, zinc-galvanized bars, and plain rebar in a marine environment [14].

Concrete specimens with different types of rebar were prepared and exposed to an inter-tidal

zone for corrosion when the age of concrete was 9 months. Visible surface staining and cracking

were observed in black steel specimens after 42 months of exposure. However, such staining was

not observed for ECR and zinc-galvanized specimens. This indicated that ECR can delay the

initialization of corrosion and provided a longer service life with little or no maintenance.

Investigations on the durability of different steel reinforcement were also conducted in the

UK. In research conducted by Swamy [15], concrete specimens with uncoated black steel, zinc-

galvanized steel, and epoxy coated steel were cast and exposed to a natural marine environment

as well as an accelerated corrosion environment. The research concluded that uncoated steel bars

corroded extensively under natural marine conditions. A concrete cover as thick as 70 mm could

not prevent corrosion. Galvanized bars showed improved performance under both marine

conditions and accelerated corrosion conditions. However, progressive corrosion was observed

under the natural environment. ECR had the best performance among the three tested bars and

survived corrosion in both marine conditions and accelerated corrosion conditions, even with a

shallow concrete cover of 20 mm. The corrosion of ECR with artificial surface damage was also

negligible and far less than uncoated and galvanized bars. The research concluded that ECR had

the best performance.

1.3.2 Corrosion Performance of ECR in Concrete Bridge Decks

Field investigation of twelve sites in Virginia showed that ECR bridge decks [16] revealed a

significant number of transverse cracks often extending through the entire depth of the deck after

10

15-20 years in service, but these cracks resulted from mechanisms other than corrosion. The use

of ECR did not prevent these cracks, but significantly reduced the corrosion-induced cracks.

Bridge decks not using ECR presented widely varying percentages of delamination from 60-80

percent, although 5-20 percent was more common. Bridge decks using ECR did not reveal much

delamination. ECR was found to be the main reason for the lack of delamination in bridge decks.

A limited field survey carried out in Minnesota [17] showed that bridges with ECR showed

no signs of distress after almost 20 years of service and no corrosion was observed on the epoxy

coated bar.

The advantage of ECR under field conditions was also verified by an investigation of eleven

bridges with bare reinforcement and eleven bridges with ECR [18]. The investigation observed

that the decks with bare reinforcement were in the initial stages of corrosion deterioration while

decks with ECR showed no signs of corrosion, indicating that ECR delayed the initiation of

corrosion.

A two year investigation on the effectiveness of ECR in concrete bridge decks indicated that

the condition of bridge decks using ECR in Pennsylvania and New York performed very well

[19]. The occurrence of progressive corrosion was less than 3 percent in Pennsylvania but

coating adhesion reduction or loss was prevalent. Half of the bridge decks exhibited adhesion

reduction of some extent within 6-10 years in service. Coating adhesion loss also occurred with

corrosion but it could not be regarded as a predictor of corrosion (i.e., if corrosion occurs there

was coating adhesion loss, but coating adhesion loss did not necessarily indicate the occurrence

of progressive corrosion). The results of this investigation were based on bridges with a low

average age (around 10 years in service).

11

The Indiana Department of Transportation performed a large scale field investigation on the

performance of ECR bridge decks [1]. One hundred twenty three bridges were visually inspected

and 44 percent indicated reinforcement corrosion. Six bridge decks (3 decks with black steel and

3 decks with ECR) were selected for detailed surveys. Corrosion of reinforcement was observed

in 2 of the 3 ECR bridge decks. Field visits to the construction site showed that the average

number of holidays, or defects, created during construction was 9 per foot. Increasing the

thickness of the epoxy coating reduced the coating damage during construction. Cracking and

insufficient concrete cover decreased the effectiveness of ECR as a corrosion protection method.

Laboratory experiments showed that cracking allowed chlorides and oxygen to reach the

reinforcement. This provided the necessary conditions for corrosion to occur. Increasing the

thickness of the concrete cover also helped prevent the penetration of chloride and oxygen and

therefore protected the reinforcement from corrosion. The reason for the better performance of

ECR was that it increased the electrical resistance dramatically. Researchers recommended that

increasing the thickness of epoxy coating and concrete cover could improve the corrosion

resistance of ECR.

Investigation on ECR bridge decks in Iowa [20] concluded that ECR bridges showed no

delamination or spalling after 20 years of use and that the concrete cover played an important

role in the initiation of corrosion. Most of the ECR corrosion was found in cores displaying

concrete cracking.

The FHWA performed a test on the long-term performance of ECR in concrete severely

contaminated by salt [21]. Bare steel had poor performance. The performance of ECR depended

on the bottom layer. If the bottom layer was black steel, the corrosion of the top layer (ECR) was

about 50 percent of that in black steel even when the coating was damaged. If the bottom layer

12

was ECR, the corrosion was less than 2 percent of that of black steel and approached the

performance of stainless steel reinforcement. Cracking reduced adhesion and coating

disbondment (or adhesion loss) was found to be the major reason for corrosion of ECR.

Brown [11] investigated bridges in Virginia and concluded that the early age cracking of

concrete decks may increase chloride penetration but did not significantly reduce the service life

of bridge decks. The use of ECR provided an extension in service life of about 5 years, making

ECR not a cost-effective corrosion protection method for bridge decks in Virginia.

Smith and Virmani [22] summarized the results of investigations conducted by different

highway agencies on the performance of ECR. A total of 92 bridge decks, 2 bridge barrier walls

and 1 noise barrier wall were investigated. ECR provided effective corrosion protection for up to

20 years of service with little or no maintenance and signs of damages related to the corrosion of

ECR were not found. The performance of ECR was not good if the concrete was cracked, the

concrete cover was small and concrete permeability and chloride concentration was high. The

defects and disbondment in the epoxy coating affected the performance of ECR.

Lee and Krauss [23] conducted a 3-year research project on the service life of ECR bridge

decks. Phase I of the research investigated 11 bridges with black steel in the bottom layer and

ECR in the top layer. Field investigation showed that bridges with ages from 19 - 27 years had

an average surface damage of 1.3 percent. Phase II of the research investigated bridges with ECR

in both the top and bottom layers. The bridges showed no sign of damage after 9-15 years of use.

A statistical model was used to predict the extension of service life brought by the use of ECR.

The results showed that bridge decks with ECR in the top layer extended the service life for

13

about 40 years or more, while bridge decks with ECR in both layers should have a service life

extension of more than 82 years.

In addition to bridge decks, ECR has been used in structures as well. An investigation of 12

garage structures [24] showed that ECR is a worthy investment if properly installed. With a few

exceptions, most of the ECR performed well. Thus, with improved quality control and

construction standards ECR was seen as a valuable corrosion control option.

The performance of ECR, however, was found to be unsatisfactory based on research

performed in Florida on marine substructures [25]. Severe corrosion of ECR was observed after

6-10 years and corrosion proceeded underneath the epoxy coating even though the coating was

unaffected. The cause of the damage was identified to be mechanical fabrication and possible

coating disbondment before and after the placement of epoxy bars in concrete. To confirm this,

laboratory experiments with and without coating defects in ECR were performed. After one-

month of exposure to corrosive agents (calcium hydroxide and sodium chloride), delaminations

around defects and pits with corrosion products were observed and low pH liquid accumulated

between the coating and the metal. Exposure of steel to neutral sodium chloride under open

circuit conditions caused significant disbondment around existing or introduced coating

imperfections, which led to the extensive corrosion in ECR.

1.3.3 Summary of the Section

This section summarized the results of previous research on the performance of ECR. The

use of ECR generally seemed to increase the corrosion resistance, although findings were mixed.

The effectiveness of the ECR is influenced by several factors such as the existence of concrete

cracks, the thickness of concrete cover, the type of steel for the bottom reinforcement, and the

14

type of the coating. Also, corrosion of ECR was observed in both laboratory environments as

well as in field investigations. Possible reasons for the corrosion of ECR are the damage of the

coating during the transportation and construction process and the disbondment of the coating

from the steel.

1.4 Repair Options for Reinforced Concrete Bridge Decks

As summarized in the previous section, although ECR can significantly increase the

corrosion resistance of bridge decks, signs of corrosion were observed in both laboratory

experiments and field investigations. Because corrosion-induced deterioration is progressive,

inspections for damage assessment must be performed routinely and repairs are needed

continually. For example, corrosion-induced concrete spalls occur as potholes in a bridge deck

and contribute to poor ride quality. In extreme conditions, structural failure and collapse may

occur [10]. Due to the difference in the corrosion mechanism and service life, the inspection and

repair schedule for ECR bridges should be different from bridge decks with black steel. However,

research on repair protocols for ECR bridge decks is limited.

In 2002, an NCHRP-sponsored project focused on investigating the repair and rehabilitation

of bridge components containing ECR. Repair options for different types of damage were

evaluated by both field investigation and laboratory tests [26]. The research concluded that

neither admixed nor migrating corrosion inhibitors provided any corrosion protection when used

to rehabilitate cracks and delaminations on concrete elements reinforced with ECR that were

exposed to an aggressive chloride environment. The research also showed that patching the

concrete using high resistivity, low permeability, silica fume concrete as well as a combination

of epoxy coating and patching provided protection against corrosion in the repair area. This

15

report also included a decision matrix for the repair of different types of damage such as

corrosion induced and non-corrosion induced cracking and corrosion induced delaminations,

based on the severity of the damage and the probability of corrosion.

The selection of repair strategies should be based on the following criteria:

1. The strategy should be expected to have a direct and positive impact on the future

performance of epoxy-coated reinforcing steel.

2. Each technique or material should not have any known adverse effects on ECR.

3. Emphasis should be placed on strategies which have a history of success on bare

reinforcing steel.

4. Each strategy should have some probability of success based on current knowledge even

if adequate long-term performance information on bare reinforcing steel is not available.

NCHRP report 558-RC “Manual on service life of corrosion-damaged reinforced concrete

bridge super-structure elements” [27] describes the commonly used repair options including

overlays, membranes, surface coatings, and sealers. Overlays can restore the surface of the

bridge deck and increase the effective thickness of concrete cover. The cost of concrete removal

and chloride concentration in the concrete below the depth of removal must be considered when

determining the extent of material to be removed. The service life of rehabilitation overlays is

limited by the rate of diffusion of chloride ions through the overlay and can be significantly

influenced by environmental exposure conditions (i.e., chloride concentration and temperature).

The membrane can provide a waterproofing barrier to prevent the intrusion of chloride ions into

the concrete deck and is usually used in conjunction with an asphalt overlay. Proper application

16

of an approved membrane can greatly reduce the intrusion of chlorides into the concrete. The

advantage of the membrane is that it can be applied relatively quickly to almost any deck

geometry and can bridge most moving concrete cracks because of its elastic nature. However, the

service life of the membrane may be limited by wearing when exposed to heavy traffic and by

the disbondment and shoving under traffic. Concrete sealers and surface coatings can also be

used to prevent chloride ions from diffusing into the concrete, but they cannot reduce the

chloride that is already in the concrete before the repair. The performance of the concrete sealer

or surface coating is also dependent on the environment, including the level of UV (ultra-violet)

radiation, moisture, and abrasion. The extension in service life by surface coating is usually

limited.

The Michigan Department of Transportation (MDOT) has a developed a repair matrix for

bridge decks with black steel reinforcement in which the repair options were selected based on

the level of damage (percentage of damaged surface) and has been successfully applying these

repair options. However, the influence of these repair options on the performance of bridge decks

with ECR needs to be evaluated.

1.5 Summary

This section first provided a brief review on the corrosion mechanism of steel reinforcement

in concrete. Different corrosion protection or corrosion control methods were described. ECR

was found to be the most cost-effective corrosion control method in most cases. Next, a review

of research on the performance of ECR in concrete bridges and other types of structures was

provided. From the review, it is apparent that ECR has better corrosion performance than black

steel reinforcement. However, some corrosion of ECR was observed both in laboratory tests and

17

in field investigations. Therefore, inspection and repair of ECR bridge decks is necessary. The

last section summarized research on the repair options for structures with ECR.

18

CHAPTER 2

1

SPECIMEN PREPARATION

Damage to bridge decks is caused by a combination of mechanical fatigue and environmental

loads. Laboratory specimens need to be conditioned to simulate the actual aging effects before

being used to evaluate service life. Bridge decks in Michigan are usually subjected to two main

types of environmental loads: freeze-thaw effects and the corrosion of steel bars due to the use of

de-icing salts during winter. Therefore, the conditioning of specimens in this project included

freeze-thaw cycles and corrosion of the steel. After the specimens were conditioned, they were

repaired by different methods according to the damage level they experienced. The conditioned

and repaired specimens were then used to evaluate the service life and performance of different

repair options.

2.1 Specimen Design and Casting Preparation

To simulate the performance of bridge decks under field conditions, large scale testing is

needed. However, cost and logistical considerations often limit the number of configurations that

can be tested, and typically only one specimen can be tested each time. In this research, small-

scale specimens with different configurations were used. To evaluate the relative performance of

bridge decks with epoxy coated rebar (ECR), specimens with black steel rebar (BSR) were used

for comparison. Each test unit was fabricated with two rebars. For specimens with black steel,

both bars in the test unit consisted of black steel. For specimens with ECR, the top bar was epoxy

coated but the bottom bar was made of black steel so that it could be used as a cathode in the

19

accelerated corrosion process. Figure 2.1 shows the sizes and rebar configuration of the proposed

specimens.

Figure 2.1: Specimen Configuration

To evaluate the performance of the specimens under different damage levels and repair

options, a representative sampling of the repair matrix used by the Michigan Department of

Transportation (MDOT) was used, as listed in Table 2.1.

The coating of the ECR will inevitably get damaged during construction. Based on previous

research and suggestions from the Research Advisory Panel at MDOT, a surface damage of 2

percent was selected. The damage was introduced by lightly drilling the epoxy coated surface of

the rebar until the epoxy under the 1/16" drill bit was removed. The location of the damage was

randomly selected. One example of the damaged rebar is shown in Figure 2.2.

#3 Epoxy-Coated or

Black Steel Rebar

#3 Black Steel

Rebar

4"

3"

16"

20

Table 2.1: Test Matrix

Test ID Damage Level Repair Option Steel Type

B0-N 0 None BSR

E0-N 0 None ECR

B1-N 1 None BSR

E1-N 1 None ECR

B1-R 1 Epoxy Overlay BSR

E1-R 1 Epoxy Overlay ECR

B2-N 2 None BSR

E2-N 2 None ECR

B2-R 2 Shallow Concrete Overlay BSR

E2-R 2 Shallow Concrete Overlay ECR

B3-N 3 None BSR

E3-N 3 None ECR

B3-R 3 HMA Overlay with Waterproofing BSR

E3-R 3 HMA Overlay with Waterproofing ECR

B4-N 4 None BSR

E4-N 4 None ECR

B4-R 4 Deep Concrete Overlay BSR

E4-R 4 Deep Concrete Overlay ECR

21

Figure 2.2: ECR with Damage in Coating

Due to the small size of the specimens, it was likely that shear failures would be

precipitated by slipping of the tension reinforcement. Once shear cracks initiated, sudden failure

would have occurred since the test units had no shear reinforcement. Three pilot test units were

thus designed, built, and tested to evaluate if the desired strain levels induced by traffic on

bridges could be reached before the onset of shear failure. The first test unit (Beam 1) was only

reinforced with two #3 bars at the top and bottom of the test unit. In the second test unit (Beam 2)

the tension bar had washers welded at both ends to provide additional mechanical anchorage.

The third beam (Beam 3) had no special anchorage detail, but had a steel wire mesh as shear

reinforcement. The dimension and reinforcement lay-outs are shown in Figure 2.3.

The beams were loaded in three-point bending to maximize flexural demands and

minimize shear demands. The total simply supported span was 16 inches. The beams were

loaded monotonically under displacement control until failure. Load, mid-span displacement,

Coating Damage

22

and strain of the tension bar at mid-span were monitored. The average concrete compressive

strength when the beams were tested was 6,790 psi.

(a) Beam 1

(b) Beam 2

(c) Beam 3

Figure 2.3: Dimensions and Reinforcement Layouts of Pilot Beams

#3 Bars

Welded Washers

Wire mesh

# 3 Bars

Welded Washers

4"

4"

4"

3"

3"

3"

16"

22"

16"

16"

22"

22"

23

The load-displacement response for all beams is shown in Figure 2.4 and photographs of

their failure pattern are shown in Figure 2.5. Beam 1 (without any anchorage enhancement)

failed before yielding of the tension rebar. Beam 2 (with welded washer on bottom bar) failed in

shear, but the specimen failed after the yielding of the bottom reinforcement. Beam 3 (with wire

mesh shear reinforcement) also failed in shear and the tension bar yielded before failure. From

the measured responses, observed failure, and crack patterns, Beam 3 performed the best with a

rather smooth response up to failure with the tension bar reaching strains well beyond yield.

Beam 2 performed the next best and was satisfactory. Since welding washers was much easier

than installing the wire mesh, washers were used for fabrication of the project test units. The

molds with the steel rebars before casting is shown in Figure 2.6.

Figure 2.4: Load-Displacement Response of Pilot Test Beams

Load_Displacement Curve

0

1000

2000

3000

4000

5000

6000

0 0.02 0.04 0.06 0.08 0.1 0.12

Disp. (in.)

Load (

lbs)

Beam 1

Beam 2

Beam 3

Beam 1

Beam 2

Beam 3

0 0.02 0.04 0.06 0.08 0.10 0.12

Disp. (in.)

Lo

ad (

lbs)

6000

5000

4000

3000

2000

1000

0

24

(a) Beam 1

(b) Beam 2

(c) Beam 3 Figure 2.5: Failure Patterns of Pilot Test Beams

25

Figure 2.6: Molds before Casting

2.2 Accelerated Freeze-Thaw Test

Damage in concrete bridge decks occurs due to a combination of mechanical and

environmental loading. All specimens were “aged” by exposing them to freeze-thaw cycling and

accelerated corrosion. The initial conditioning imposed was equivalent to approximately 30 years

of aging, which corresponded to the approximate age of current ECR decks in Michigan. Based

on 3 years of temperature data in the Lansing area, 300 freeze-thaw cycles were used to simulate

30 years of aging. The specimens were subjected to freeze-thaw cycling in MDOT’s freeze-thaw

machine. ASTM C-666 [28] Procedure B was used in the test. Due to an incorrect mixture design,

the first batch of specimens experienced severe damage after the freeze-thaw test as shown in

Figure 2.7(a), and could not be used for further testing. The reason for the severe damage was

later found to be insufficient air-entrainment in the concrete. The mix design was revised and a

second batch of concrete with a measured air-content of 6.1 percent was cast. An average

26

compressive strength of 4.3 ksi was achieved after 28 days of curing. The specimens from the

second batch successfully survived the freeze-thaw test as shown in Figure 2.7(b).

(a) First Batch

(b) Second Batch

Figure 2.7: Specimens after Freeze-Thaw Test

27

2.3 Accelerated Corrosion

In addition to freeze-thaw, bridge decks in Michigan are also subjected to corrosion induced

by the use of de-icing salt. Similar to freeze-thaw, corrosion occurs over a long period of time. In

this research, accelerated corrosion using an impressed current was used to simulate field effects

after the freeze-thaw test. The top bar was used as an anode and the bottom bar as a cathode. The

induced corrosion level was similar to that experienced after 30 years of service and was

estimated by monitoring the current through the specimen. The total amount of charge for

simulated corrosion is given by:

Q=∆m·z·F

Am (2.1)

where m∆ is the mass loss for a given period of time estimated from field investigations, z is the

valency charge of the metal (z = 2 for Fe2+

), F = 26.8015 Ah is Faraday’s constant, and Am is the

atomic weight of the metal (55.85 for steel).

In this project, a corrosion rate of 0.045 mm/year was selected based on previous research

[29]. The mass loss over n years can then be calculated as:

∆m = π

4R2-R - 0.045n2ρL (2.2)

where R is the average radius of the rebar, n is the number years, ρ is the density of the steel,

and L is the length of the rebar.

28

The impressed current was measured by a voltage data-logging system (Omega Engineering

Inc., AD128-10T2). Three specimens were connected in series and were connected to a 36 V DC

power supply to reduce the time of conducting replicate tests. During the accelerated corrosion

process the test units were soaked for one hour each day in a 3.5 percent NaCl solution by weight.

The circuit used for the corrosion test is shown in Figure 2.8.

Figure 2.8: Accelerated Corrosion Test Setup

Due to the small size of the specimen, the damage due to corrosion could not be effectively

measured in terms of the percent damage to the surface area. Instead, the measured corrosion

charge was used as a measure of the damage level. The maximum level of damage (Level 4) was

determined to be equivalent to 30 years of field corrosion. The corrosion charge for Level 4 can

be calculated using Equation 2.1. The minimum damage level (Level 1) was defined as the

charge when the initial crack appears in the specimen and the charge for Level 1 was obtained

Data Logging System

R=12 Ω

R=12 Ω

36 V DC

29

through observations during the test. Damage levels 2 and 3 were defined by linear interpolation

between the charges Q1 and Q4, corresponding to Level 1 and Level 4 damage. The charge for

different levels of damage is shown in Table 2.2.

Table 2.2: Corrosion Charge for Different Damage Levels

Damage Level 1 2 3 4

Corrosion Charge

(Ah) 12.57 43.65 74.72 105.79

Figure 2.9 shows an example of the corrosion curve for BSR and ECR specimens. The

corrosion rate was much slower for ECR than BSR and the time for ECR specimens to reach the

same level of corrosion was approximately 2-3 times longer than for BSR specimens. Therefore,

ECR seems to effectively delay the appearance of the corrosion induced damage in concrete

bridge decks.

Figure 2.9: Corrosion Curves for BSR and ECR

0

20

40

60

80

100

120

0 40 80 120 160Time (Days)

Ch

arg

e(A

h)

BSRECR

Q1

Q2

Q3

Q4

0 40 80 120 160

Time (days)

120

100

80

60

40

20

0

Ch

arg

e (A

h)

Q4

Q3

Q2

Q1

BSR ECR

30

Figure 2.10 to Figure 2.13 show typical examples of specimens corroded to different levels.

Fine cracks were observed in specimens with Level 1 damage and a marker was used to

“highlight” the crack in Figure 2.10. It is obvious that the damage increased as the corrosion

level increases, as shown in the following figures.

Figure 2.10: Specimen after Level 1 Corrosion


31



2.4 Specimen Repair

After being corroded to the desired level, specimens were repaired using different options

according to the levels of damage. Specimens with Level 1 damage were repaired using an epoxy

overlay. According to MDOT Special Provision for Thin Epoxy Polymer Bridge Deck Overlay

32

[30], epoxy produced by E-bond Epoxies, Inc. was used. The fine aggregate (sand) used in this

project was #612 Quartz sand from Best Sand Inc, in Ohio. The surface of the specimen was

first roughened by chiseling off the concrete to increase the bond with the applied overlay.

According to MDOT’s provision, two courses were applied. The epoxy mixture for the first

course was 40 ft2/gal and the second course was 20 ft

2/gal. Sand was applied such that the

surface was covered in excess with no bleed or visible wet spots. A specimen repaired with

epoxy overlay is shown in Figure 2.14.

Figure 2.14: Specimen Repaired with Epoxy Overlay

Specimens corroded to Level 2 were repaired with a shallow concrete overlay. When applied

on the bridge deck, the concrete cover is removed until the top layer of the reinforcement and the

surface is roughened to provide enough bond between the deck and the newly applied overlay.

During preparation for the shallow concrete overlay for the test specimens, it was found that the

33

corrosion cracks developed from the top to the bottom bars. Due to the small size of the

specimens, it was easy to break them along the crack and damage them if a chisel was used to

roughen the surface (one specimen was damaged due to the impact). To prevent specimen

damage, a sander was used to create a rough texture on the surface to provide bonding between

the specimen and the overlay. Also due to the penetration of the crack toward the bottom bar (in

tension), specimens were prone to be broken in half during the fatigue test. To prevent this, the

concrete overlay was applied on both sides and on top of the specimen as shown in Figure 2.15.

The confinement provided by concrete on both sides helped prevent the specimen from splitting.

Moreover, this over-size repair provided confinement to the damaged specimen and reduced the

chance of delamination. The concrete mixture design and material section followed the 2003

Standard Specifications for Construction from MDOT [31]. A typical specimen after repair is

shown in Figure 2.16.

Figure 2.15: Oversized Shallow Concrete Overlay

Removed

Concrete

Crack

Shallow Concrete

Overlay

Roughened Surface

5

8"

1

2"

1

2"

3"

4"

34

Figure 2.16: Specimen with Shallow Concrete Overlay

Specimens with Level 3 damage were repaired with a hot mixture asphalt (HMA) overlay

and a water-proofing membrane. The HMA mixture and the water proofing were provided by

MDOT. The waterproofing was applied according to the 2003 Standard Specifications for

Construction [31] provided by MDOT and the asphalt overlay was applied according to the Road

Design Manual [32] from MDOT. A typical repaired specimen is shown in Figure 2.17.

35

Figure 2.17: Specimen with Waterproofing and HMA Overlay

A deep concrete overlay was used to repair specimens with Level 4 damage. Similar to the

shallow concrete overlay, an oversize repair method was used due to the small size of the

specimens. The mixture design was the same as that used in the shallow concrete overlay. Figure

2.18 shows the size of the deep concrete overlay and Figure 2.19 shows a typical specimen after

repair.

Figure 2.18: Oversized Deep Concrete Overlay

Removed

Concrete

Roughened Surface

Crack

Deep Concrete Overlay

1

2"

1

2" 4"

3"

11

2"

5

8"

36

Figure 2.19: Specimen with Deep Concrete Overlay

37

CHAPTER 3

EXPERIMENTAL DESIGN

An experiment was conducted to determine the service-life performance of concrete bridge

decks reinforced with black and epoxy coated steel. This was accomplished by testing small-

scale concrete specimens with varying amounts of damage and different repair options for each

type of steel. The specimens were subjected simultaneously to cyclic mechanical loads,

freeze/thaw cycling, and accelerated corrosion. Periodically, the testing was halted so that the

stiffness of the specimens could be measured. The stiffness values obtained throughout the

cycling process were used to evaluate the service-life performance of the specimens. These

values were used to compare the service-life performance of specimens with black steel and with

epoxy coated steel. This chapter provides the detailed experimental design used to accomplish

this goal.

3.1 Fatigue Testing

The specimens prepared for this experiment were subjected to loading that simulated the

demands that occur on a typical reinforced concrete bridge deck. The types of loads considered

were vehicular traffic, freeze-thaw cycling, and corrosion. The service-life of each specimen was

determined by simultaneously subjecting the specimens to these loads. In order to reduce the

time of this experiment, three specimens were tested at a time in an environmental chamber

which is shown in Figure 3.1.

38

Figure 3.1: Environmental Chamber

Typically, the three specimens being tested at a particular time had the same age and the

same method of repair. Since some of the specimens were damaged during the aging process,

this was not always the case. Therefore, specimens were placed into groups of three, in which the

specimens in each group had similar ages and methods of repair. A list showing the groups is

given in Table 3.1.

39

Table 3.1: List of Groups of Specimens

Group 1 B-0-1 B-0-2 B-0-3

Group 2 E-0-1 B-1-N-1 E-0-2

Group 3 B-1-R-1 B-2-R-1 B-1-R-2

Group 4 B-2-N-1 B-1-N-2 B-2-N-2

Group 5 E-1-N-1 E-1-N-2 E-1-N-3

Group 6 E-1-R-1 E-1-R-2 E-1-R-3

Group 7 E-2-N-1 E-2-N-2 E-2-N-3

Group 8 B-4-N-1 B-4-N-2 B-4-N-3

Group 9 E-4-N-1 E-4-N-2 E-4-N-3

Group 10 B-3-N-1 B-3-N-2 B-3-N-3

Group 11 E-3-N-1 E-3-N-2 E-3-N-3

Group 12 E-2-R-1 E-2-R-2 E-2-R-3

Group 13 B-3-R-1 B-3-R-2 B-3-R-3

Group 14 E-4-R-1 E-4-R-2 E-4-R-3

Group 15 E-3-R-1

E-3-R-2

Group 16 B-4-R-1 B-4-R-2 B-4-R-3

Figure 3.2: Fatigue Test Setup

Roller

Rubber Pad Support Frame

Loading Frame

40

The vehicular traffic was simulated on the specimens by cyclic flexural loading in a three-

point bending setup as shown in Figure 3.2. Rubber pads were placed between the rollers and the

specimens in order to ensure even distribution of load throughout the width of the specimens.

Two-million cycles of fatigue loading were applied to each specimen throughout this experiment

at a rate of 1 Hz. The fatigue loading applied to each group of specimens per cycle varied

between 2 kN and 14 kN, therefore applying loads varying from approximately 0.7 kN to 4.7 kN

per specimen. A minimum load of 2 kN was maintained in order to ensure that there was no

separation between the specimens and loading frame. The maximum load of 14 kN was chosen

in order to induce a tensile stress of about 0.45fy in the tension reinforcement. This limit was

determined by factoring the upper service steel reinforcement tensile stress limit of 0.6fy by 0.75

for fatigue loading [33], considering that this is what most likely contributes to concrete damage

from mechanical loads.

Freeze-thaw cycling was also applied to the specimens. While the fatigue testing was taking

place in the environmental chamber, the specimens were wetted using a water drip system. At

the same time, the temperature inside the chamber was varied from -3°F to 50°F, thus subjecting

the specimens to freeze-thaw cycles. The freeze-thaw cycles took approximately 3-4 hours per

cycle. For each specimen, the target number of freeze-thaw cycles was 150. However, the

variation in the length of the cycles caused the actual number of freeze-thaw cycles to range

between 95 and 146.

The third type of loading applied to the specimens was accelerated corrosion. This was

accomplished by impressing current through the steel reinforcement in the specimens, which

simulates the corrosion process that occurs in reinforced concrete bridge decks. The setup used

41

to accomplish this task is shown in Figure 3.3. The applied current was monitored by a voltage

data-logging unit and the total corrosion was estimated from these measurements using

Faraday’s Law.

Figure 3.3: Corrosion Setup

3.2 Static Testing

The service-life performance of the specimens was determined by obtaining the stiffness of

the specimens at certain times throughout the fatigue testing. This was done by performing static

12 Ω Resistor

36 V Power Supply

Voltage Data-Logging Unit

Wires for impressing current

42

tests during which the load-displacement response of the specimen was measured. The stiffness

was then determined by calculating the slope of the resulting load-displacement curve, which

will be discussed later.

Periodically, the fatigue test was halted so that the static tests could be performed. These tests

were conducted in the same environmental chamber as the fatigue testing, shown in Figure 3.4.

Load cells were placed between the specimens and the loading frame to measure the loads

applied to each specimen. Linear variable differential transducers (LVDTs) were mounted on the

support frame under each specimen to measure their displacements under the load. Small glass

plates were glued to the bottoms of the specimens where the LVDTs came into contact with them

as shown in Figure 3.5. This was done in order to provide a smooth surface for the LVDTs to

rest against so the roughness of the specimens would not interfere with the displacement

measurements. The load cells and LVDTs were connected to a computer so that the load and

deflection measurements could be recorded. The group of specimens was then subjected to a

load in a three-point bending setup that increased from 0 kN to 17 kN at a rate of 4 kN/min.

During this loading, data from the load cells and the LVDTs was recorded by the computer. The

result of each test was the load-deflection behavior for each of the three specimens.

43

Figure 3.4: Static Test Setup

Figure 3.5: Smooth Surface for LVDT

Loading Frame

Roller

LVDT

Support Frame

Load Cell Rubber Pad

Glass Plate

44

In order to obtain more accurate results, multiple static tests were performed. The first static

test was done only to preload the rubber pads, such that the subsequent tests would be performed

with the rubber pads having approximately the same amount of deformation. Three static tests

were then performed in succession, each one as described before. As shown in Figure 3.6, this

resulted in more consistent data, with the latter three tests showing similar load-displacement

responses. The nonlinearity in the load-displacement responses during the early stages of loading

was due to the nonlinear compressive behavior of the rubber pads used between the rollers and

the specimens. The stiffness values obtained from the curves from the latter three tests were

averaged for each specimen, obtaining a stiffness value for the specimen after a certain number

of fatigue cycles. However, as shown in Figure 3.7, the point where the load was applied and the

point where the deflection was measured were not the same. The deflection under the load was

estimated by increasing the measured deflection by seven percent as shown in the following

calculation:

For a simply supported beam: δ(x) = Px 48EI⁄ (3L2 - 4x2)

where:

δ = deflection in the specimen

L = length of specimen

x = distance from edge of specimen

δ7.5

δ5.9=

P·7.5 48EI⁄ 3·152- 4·7.52

P·5.9 48EI⁄ 3·152- 4·5.92 ≈ 1.07

45

Figure 3.6: Data Obtained from Multiple Static Tests

Figure 3.7: Schematic of Static Test

As shown in Figure 3.8, the typical load-displacement graph is non-linear. This is primarily

due to the use of the rubber pads between the roller and the specimen, as shown in Figure 3.4.

Because of this nonlinearity, the stiffness of the specimens was measured as the slope of the

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0 0.01 0.02 0.03 0.04

Load

(k

ips)

Displacement (in)

Static Test 1

Static Test 2

Static Test 3

Static Test 4

Point of Loading

Point of Measured

Deflection

x 15"

1.6"

46

load-displacement curve when the load was between 0.8 and 1.0 kips. In this range of loading,

the rubber pads were significantly compressed and therefore they were not a major factor in the

stiffness calculation.

Figure 3.8: Typical Load-Displacement Graph

0

0.2

0.4

0.6

0.8

1

1.2

1.4

0 0.005 0.01 0.015 0.02 0.025 0.03

Load

(k

ips)

Displacement (in)

Specimen Stiffness

47

CHAPTER 4

EXPERIMENTAL TEST RESULTS

The procedures for fatigue and static testing described in Chapter 3 were performed on all

repaired and non-repaired specimens and each specimen’s stiffness was estimated at specific

times during the testing. The expected result was that the stiffness of each specimen would

decrease as the fatigue testing progressed. It was envisaged that appropriate repair strategies for

ECR decks could be recommended by comparing the stiffness of specimens with black steel with

a particular repair option to the stiffness of specimens with ECR with the same repair option.

4.1 Static Test Results

Figures 4.1 to 4.8 show the average stiffness values for each group normalized relative to the

initial stiffness measurement for the same group. Figures are not shown for Level 3 unrepaired

groups because these specimens failed quickly. The black steel Level 3 unrepaired specimens

failed at 0 cycles and the epoxy coated steel Level 3 unrepaired specimens failed at 414,000

cycles. Actual stiffness values are given in Appendix A.

48

Figure 4.1: Normalized Stiffness Values (Level 0 Unrepaired)


0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

0 500000 1000000 1500000 2000000

Norm

ali

zed

Sti

ffn

ess

Cycles

Black Steel

Epoxy Coated Steel

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

0 500000 1000000 1500000 2000000

Norm

ali

zed

Sti

ffn

ess

Cycles

Black Steel

Epoxy Coated Steel

1 specimen failed at 645,000 cycles

49

Figure 4.3: Normalized Stiffness Values (Level 1 Repaired)


0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

0 500000 1000000 1500000 2000000

Norm

ali

zed

Sti

ffn

ess

Cycles

Black Steel

Epoxy Coated Steel

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

0 500000 1000000 1500000 2000000

Norm

ali

zed

Sti

ffn

ess

Cycles

Black Steel

Epoxy Coated SteelAll specimens failed

at 745,000 cycles

50



0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

0 500000 1000000 1500000 2000000

Norm

ali

zed

Sti

ffn

ess

Cycles

Black Steel

Epoxy Coated Steel

c

1 specimen failed at 1,400,000 cycles1 specimen failed at 1,404,000 cycles

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

0 500000 1000000 1500000 2000000

Norm

ali

zed

Sti

ffn

ess

Cycles

Black Steel

Epoxy Coated Steel

c

1 specimen failed at 0 cycles (black steel)1 specimen failed at 1,000,000 cycles1 specimen failed at 1,600,000 cycles

51



0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

0 500000 1000000 1500000 2000000

Norm

ali

zed

Sti

ffn

ess

Cycles

Black Steel

Epoxy Coated Steel

1 specimen failed at 0 cycles2 specimens failed at 1,600,000 cyclesAll black steel specimens failed at 0 cycles

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

0 500000 1000000 1500000 2000000

Norm

ali

zed

Sti

ffn

ess

Cycles

Black Steel

Epoxy Coated Steel

2 specimens failed at 0 cycles (epoxy coated steel)

52

The stiffness values obtained from the static tests did not show a consistent decreasing trend

throughout the fatigue testing as expected. Generally, the stiffness of a given specimen did not

change significantly and in many cases, stiffness values increased. Therefore, the stiffness

measurements were considered unreliable and not particularly useful.

Some comparisons of the performance between black and epoxy coated steel can be made by

examining when some of the specimens failed. Table 4.1 shows the number of cycles to failure

for each test specimen. Some specimens were damaged during the repair procedure and hence

three replicates were not available for each case. For unrepaired specimens with corrosion levels

0-3, those with black steel always failed before those with epoxy coated steel. Unrepaired

specimens with corrosion level 4 all failed when first loaded. Repaired specimens had a more

haphazard behavior. In general, epoxy coating helped improve survivability under fatigue

loading.

Table 4.1: Survival Under Fatigue Loading

Repair Type/Corrosion Level Number of Cycles to Failure*

Black Steel ECR

Unrepaired/Level 0 −; −; − −; −

Unrepaired/Level 1 −; 644,222 −; −; −

Unrepaired/Level 2 744,710; 744,710 −; −; −

Unrepaired/Level 3 0; 0; 0 414,523; 0; 414,523

Unrepaired/Level 4 0; 0; 0 0; 0; 0

Epoxy Overlay/Level 1 −; − −; −; −

Shallow Concrete Overlay/Level 2

− −; 1,436,412; 1,401,185

HMA Overlay/Level 3 −; 0; 1,053,614 −; 1,591,507

Deep Concrete Overlay/Level 4

−; −; − 0; 0; −

* The symbol − indicates that a specimen did not fail during the 2 million fatigue cycles

53

4.2 Corrosion during Fatigue Loading

During fatigue testing the specimens were subjected to accelerated corrosion. This was done

by impressing current through the specimens as described in Sections 2.3 and 3.1. The impressed

current was measured throughout the fatigue testing, which was then used as a measure of the

corrosion-induced damage in the specimens. A typical graph showing the corrosion charge vs.

fatigue cycles is shown in Figure 4.9.

The corrosion rate was slower for epoxy coated steel than for black steel by a factor of about

2.5. Therefore, similar to the observations in Section 2.3, epoxy coated steel should effectively

delay the appearance of corrosion induced damage compared to black steel. The corrosion data is

discussed in more detail in Chapter 6.

Figure 4.9: Charge vs. Fatigue Cycles

0

2

4

6

8

10

12

14

16

18

0 500000 1000000 1500000 2000000

Ch

arg

e (A

h)

Cycles

Black Steel

Epoxy Coated Steel

54

CHAPTER 5

X-RAY COMPUTED TOMOGRAPHY

As discussed in Chapter 4, the stiffness values obtained during the fatigue testing did not

provide results that were useful in quantifying the damage done to the concrete specimens with

black steel and epoxy coated steel. X-ray tomography had the potential to identify internal

damage in concrete specimens. The concrete specimens were scanned with X-rays, resulting in

images at specific cross-sections of the specimens. These images showed the cracks inside the

specimens, which were used as an indication of the amount of damage done during the fatigue

testing.

5.1 X-Ray Scanning

The specimens were scanned at the Center for Quantitative X-Ray Imaging (CQI) at Penn

State University [34]. This facility houses the HD600 (OMNI-X) industrial high-resolution X-ray

computed tomography (CT) scanner, which was used to scan the specimens [35]. Figure 5.1

shows a photograph of the setup used to do the scanning. A specimen was placed in a tube of

sand and rotated on a plate as it was scanned 2400 times in order to produce the best quality

image possible at a particular location. Eighty-nine images along the length of the specimen,

each with a separation of 0.0046 inches, were taken at four locations in each specimen. Two

locations were 1 inch on either side of the mid-length, and the other two locations were 2 inches

on either side of the mid-length. Therefore, at each location, images of the cross-sections of the

specimen were captured along a 0.409 inch length of the specimen. These images were used to

determine the volume of cracks within the four 0.409 inch long segments of each specimen. An

example of an image is shown in

5.3 for comparison.

Figure

Figure

55

example of an image is shown in Figure 5.2 with a photograph of the specimen shown in

Figure 5.1: X-Ray Scanning Setup

Figure 5.2: Example of X-Ray Scan

with a photograph of the specimen shown in Figure

56

Figure 5.3: Photograph of Scanned Specimen

This process was conducted for each specimen that did not break apart during the fatigue

testing. A total of 28 specimens were scanned. Images that show a typical scan and its

corresponding threshold image from each repair type are shown in Appendix B.

5.2 Image Processing

The main goal of the X-ray scans was to quantify the damage done to each specimen as a

result of corrosion, freeze-thaw, and fatigue loading. This was done by computing the volume of

the cracks that were visible in the X-ray images. As described in Section 5.1, each specimen had

four locations scanned and at each location a thickness of 0.409 in. was scanned 89 times,

allowing for the volume of cracks to be calculated. Software was used to compile the images and

calculate the volume of cracks [36]. However, before this software was used, a number of steps

were performed to process the images so they could be used. The image from Figure 5.2 will be

57

used as an example to illustrate the process. The software used for the following steps is called

“ImageJ” [37].

1. Crop the desired crack from the rest of the image.

Figure 5.4: Crack Cropped from Image

2. Invert the colors.

Figure 5.5: Crack with Inverted Colors

58

3. Adjust contrast and brightness. The values used to adjust these parameters varied

between specimens such that the crack was easily shown in each case.

Figure 5.6: Crack with Enhanced Brightness and Contrast

4. Apply a bandpass filter to reduce noise.

Figure 5.7: Crack after Applying Bandpass Filter

59

The images were then imported into the XCAT software and a threshold was applied, converting

the grayscale images into black and white.

Figure 5.8: Crack after Applying Threshold

This process was done for all 89 images at each location. The XCAT software reconstructed

the images at each location and created a 3D rendering of the crack, as shown in Figure 5.9. The

software then calculated the volume of the crack in voxels (volumetric pixel). The total volume

of cracks in a specimen was used as a measure of the damage done to it by corrosion, freeze-

thaw, and fatigue testing. The results showing the image-based damage for each specimen

scanned are shown in Table 5.1 along with the time that each specimen was exposed to

accelerated corrosion (initial corrosion for aging plus corrosion during fatigue loading).

60

Figure 5.9: 3D Rendering of Crack

Table 5.1: Image-Based Damage

Specimen

Image-Based

Damage

(voxels)

Time

(days) Specimen

Image-Based

Damage

(voxels)

Time

(days)

B-0-1 0 23 E-0-1 0 23

B-0-2 0 23 E-0-2 0 23

B-0-3 89,285 23 E-1-N-1 0 44

B-1-N-1 18,733 35 E-1-N-2 0 44

B-1-N-2 210,555 35 E-1-N-3 0 44

B-1-R-1 1,850,790 35 E-1-R-1 0 67

B-1-R-2 105,660 35 E-1-R-2 0 67

B-2-R-2 1,009,233 54 E-1-R-3 0 67

B-3-N-2 520,482 62 E-2-N-2 0 147

B-3-R-2 1,951,323 79 E-2-N-3 0 147

B-4-R-1 1,165,097 86 E-2-R-2 373,500 144

B-4-R-2 0 86 E-3-N-1 757,362 175

B-4-R-3 754,504 86 E-3-N-3 126,751 175

E-3-R-2 307,464 144

E-4-R-2 15,311 182

61

CHAPTER 6

INTERPRETATION OF RESULTS AND CONCLUSIONS

The goal of this part of the project was to determine the maintenance schedule for ECR

concrete bridge decks. The Michigan Department of Transportation (MDOT) has an established

maintenance schedule for concrete bridge decks with black steel based on the amount of surface

and underside damage on decks. By comparing the times to equivalent damage levels in ECR

and black steel concrete decks, the maintenance schedule for black steel decks could be revised

for ECR decks. Damage levels were defined in two ways: (1) by using the corrosion charge

impressed during specimen aging described in Section 2.3 (Phase 1) and during the fatigue

testing described in Section 4.2 (Phase 2); and (2) by comparing the image-based damage

between specimens as described in Section 5.2.

6.1 Corrosion Damage

The cumulative corrosion charge measured during the test was used as a measure of how

much the steel corroded as evidenced by the physical appearance (staining, cracking) of the

specimen. In other words, how much total charge in Ah was impressed to the specimen at the

time of damage observation. A comparison of the corrosion charge for epoxy coated and black

steel reinforcement would reveal the relative improvement against corrosion provided by the

epoxy coating. Figure 6.1 shows a plot of the corrosion charge as a function of the time for

which each specimen experienced accelerated corrosion in the laboratory. Data throughout Phase

1 and Phase 2 are plotted in this figure for all specimens, including repaired and unrepaired. The

data displayed in this figure is shown in Appendix C. The approximate field time scale is also

62

shown on the top (x-axis). This scale was determined by utilizing the fact that a certain amount

of charge corresponds to a certain amount of time in the field as described in Section 2.3. The

values from Table 2.2 were used where Level 4 damage corresponded to 30 years in the field and

Level 1, 2, and 3 damage levels corresponded to 7.5, 15, and 22.5 years in the field, respectively.

The time for each group of black steel specimens to reach each level of damage when exposed to

accelerated corrosion was averaged and is shown in Table 6.1. This data suggests that the

relationship between the time of accelerated corrosion and the corresponding time in the field is

linearly related according to Equation 6.1 where tfield is the time in the field and tac is the time

exposed to accelerated corrosion (both expressed in the same units).

tfield=181.7tac (6.1)

Table 6.1: Estimates of Time in the Field for Black Steel

Time in the

Field (years)

Target Corrosion

Charge (Ah)

Average Time of Accelerated

Corrosion (days)

7.5 12.57 12.8

15 43.65 29.7

22.5 74.72 45.7

30 105.79 60.6

The rate of corrosion charge is higher in black steel than in epoxy coated steel. The data

displays considerable scatter, but the type of repair does not seem to have a significant impact on

the rate of corrosion during Phase 2. This may be because during fatigue testing moisture could

enter the specimens from the sides even though the repair may affect the permeability through

the top surface. The behavior may be different in the field where moisture can only penetrate

through the top surface.

63

Figure 6.1: Charge vs. Time for Phase 1 and Phase 2

The scatter in Figure 6.1 is non-uniform and the variance increases with time (i.e., scatter is

more at larger times). Regression lines fitted to the data should consider the non-constant

variance and a weighted least squares fit should be used [38]. Regression lines based on both

regular and weighted least squares analysis were fitted for the black steel and the epoxy coated

steel and are shown in Figure 6.1. The regular least squares regression lines are shown as solid

lines and the weighted least squares regression lines are shown as dotted lines. The difference

between the two sets of lines is not large.

The slopes of these lines (i.e., rate of corrosion charge) were used to compare the corrosion

rates of the steel bars. Based on the weighted least squares regression lines, the rate of corrosion

in epoxy coated steel was about 2.6 times slower than in black steel (i.e., 1.291/0.495 = 2.60). If

y = 1.367x

y = 0.545x

y = 1.291x

y = 0.495x

0 10 20 30 40 50 60 70 80 90 100

0

20

40

60

80

100

120

140

0 20 40 60 80 100 120 140 160 180 200

Time in the Field (years)C

harg

e (A

h)

Time (days)

Phase 1 - BSR (solid) Phase 2 - BSR (hollow) Phase 1 - ECR (solid) Phase 2 - ECR (hollow) Regular LS Weighted LS Unrepaired Epoxy Overlay Shallow Concrete Overlay HMA Overlay Deep Concrete Overlay

64

-3

-2

-1

0

1

2

3

0 50 100 150 200

Wei

gh

ted

Res

idu

als

Time (days)

-40

-30

-20

-10

0

10

20

30

40

50

0 50 100 150 200

Res

idu

als

Time (days)

-40

-30

-20

-10

0

10

20

30

40

50

0 25 50 75 100

Res

idu

al

Time (days)

-3

-2

-1

0

1

2

3

0 25 50 75 100

Wei

gh

ted

Res

idu

al

Time (days)

the regular least squares regression lines are used, then the rate of corrosion in epoxy coated steel

was about 2.5 times slower than in black steel (i.e., 1.367/0.545 = 2.51).

The residuals (measured charge – charge predicted by the regression line) and weighted

residuals (residual multiplied by the weight used for each point) for the black steel and ECR are

shown in Figure 6.2 and Figure 6.3, respectively. The residuals are heteroscedastic while the

weighted residuals are reasonably homoscedastic, indicating that the weights used were

appropriate. There were also no significant outliers in the data.

Figure 6.2: Residual and Weighted Residual Plots for Black Steel

Figure 6.3: Residual and Weighted Residual Plots for ECR

65

Figure 6.3 shows that the residuals for ECR are mostly negative toward the beginning of the

accelerated corrosion testing when using a simple linear regression model. The data in the early

stage was examined to see if a better fit could be obtained. Figure 6.4 shows the data in the early

stage of accelerated corrosion for ECR. This data seems to indicate a slow progression in the

corrosion rate for approximately 12 days (equivalent to about 6 years in the field) after which a

higher corrosion rate ensues. A bilinear regression model was used to capture this behavior and

provides a better fit for ECR. However, black steel did not show this trend. The bilinear fit

indicates that for the first 12 days (6 years in the field) the rate of corrosion in ECR is about 13

times slower than that of black steel. After this the rate of corrosion in ECR is about 2.2 times

slower than that of black steel.

Figure 6.4: Early Stage of Accelerated Corrosion for ECR

Linear Fity = 0.545x

Bilinear Fity = 0.1021x

Bilinear Fity = 0.6073x - 6.3174

0 5 10 15 20 25

0

5

10

15

20

25

30

0 10 20 30 40 50

Time in the Field (years)

Ch

arg

e (A

h)

Time (days)

Phase 1 - ECR (solid) Phase 2 - ECR (hollow) Unrepaired Epoxy Overlay Shallow Concrete Overlay HMA Overlay Deep Concrete Overlay

66

6.2 Image-Based Damage Results

The volumes of cracks in the specimens, as described in Section 5.2, were used to assess the

damage done to the specimens during Phase 1 and Phase 2. All of the cracks visible in the X-ray

images of the specimens were assumed to be due to corrosion only and not from the fatigue

loading. The fatigue loading induced flexural cracks, which closed when the specimens were

unloaded and were not visible in the X-ray images. The data in Table 5.1 was plotted against the

time each specimen was subjected to the accelerated corrosion in Phase 1 and Phase 2. This plot

is shown in Figure 6.5, with the image-based damage normalized by dividing by 2,000,000

pixels.

The data from the deep concrete overlay specimens were removed from this plot. In the deep

concrete overlay repairs, the damage in the concrete that was removed (upper half of the

specimen) was completely repaired and a new ECR bar was used, and therefore the image-based

damage data was skewed for that group of specimens. Other than for the deep concrete overlay,

the repairs did not significantly affect the corrosion induced cracking.

Although the scatter is quite large, and the number of data points is small, since many

specimens did not survive the fatigue testing, polynomial curves were fitted for each type of steel.

The residual plot when all points were included in the fit is shown in Figure 6.6. The one outlier

identified was excluded when obtaining the fit shown in Figure 6.5. These curves were used to

compare the “average” damage in specimens with epoxy coated and black steel.

The times predicted by the curves for attainment of four arbitrary damage levels are shown in

Table 6.2. A time scale factor was computed by dividing the time to attain a given damage level

for epoxy coated steel by the corresponding time for black steel. For normalized levels of

67

damage up to 0.2, specimens with epoxy coated steel took about four times longer to attain the

same damage state as specimens with black steel.

Figure 6.5: Image-Based Damage vs. Time

Table 6.2: Time Scale Factors from Image-Based Damage

Image-Based

Damage

Time (days) Time Scale

Factor Black ECR

0.05 29.2 113.9 3.90

0.1 35.9 140.9 3.92

0.15 40.5 162.4 4.01

0.2 44.3 180.7 4.08

y = 0.0003x2 - 0.0121x + 0.1474

R² = 0.8953

y = 1E-05x2 - 0.0007x

R² = 0.4612

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

0 50 100 150 200

Norm

ali

zed

Im

age-

Base

d D

am

age

Time (days)

Phase 1 - BSR (solid) Phase 2 - BSR (hollow) Phase 1 - ECR (solid) Phase 2 - ECR (hollow)

Unrepaired Epoxy Overlay Shallow Concrete Overlay HMA Overlay Deep Concrete Overlay

Outlier excluded from fit

68

Figure 6.6: Residual Plot with All Points Included

6.3 Summary and Conclusions

Concrete beam specimens were fabricated with an epoxy coated or black steel top bar and a

black steel bottom bar. The specimens were aged by subjecting them to 300 freeze-thaw cycles

and then accelerated corrosion using an impressed current. Different groups of specimens were

subjected to four corrosion levels and four different types of repairs were performed. Level 4

corrosion was taken to be the corrosion experienced over 30 years in the field assuming a

reduction in diameter of 0.045 mm/year. Level 1 corrosion was the charge required to induce the

first cracks in the specimens. Levels 2 and 3 had corrosion charges equally spaced between the

charges required for Levels 1 and 4. Specimens corroded to levels 1, 2, 3, and 4 were repaired

with an epoxy overlay, shallow concrete overlay, asphalt overlay with a waterproofing

-0.4

-0.2

0

0.2

0.4

0.6

0.8

0 20 40 60 80 100

Res

idu

al

Time (days)

Outlier

69

membrane, and a deep concrete overlay, respectively. The repaired specimens were then

subjected to fatigue loading using 3-point bending for 2 million cycles or until failure, combined

with freeze-thaw exposure and continued accelerated corrosion. Control specimens that were

aged and not repaired and other control specimens that were not aged or repaired were also

subjected to the combined fatigue loading, freeze thaw exposure, and accelerated corrosion.

Static flexural tests were performed periodically throughout the fatigue loading to measure the

flexural stiffness of the beam specimens.

X-ray computed tomography (CT) scans of the specimens that survived the 2 million cycles

of fatigue loading were performed. The volume of corrosion-induced cracks at four cross

sections, each having a width of 0.409 inches, were computed from the X-ray CT images and

used as another measure of damage.

The following conclusions are made based on the test data obtained in this project:

• The impact of damage growth due to corrosion, fatigue cycles at service load levels,

and freeze-thaw cycling on the flexural stiffness of repaired and unrepaired beam

specimens does not show any consistent trends for specimens with epoxy coated and

black steel reinforcement. Stiffness measurements were therefore not useful in

quantifying damage.

• The electrical charge impressed to concrete specimens in accelerated corrosion testing

indicated that epoxy coated steel with 2 percent surface damage to the coating

corroded at a rate that was about 2.5 times slower than black steel. The corrosion rate

was not affected significantly by different repairs performed to the concrete surface

after initial aging. Fatigue loading also did not have an effect on the corrosion rate.

70

• The volume of corrosion-induced cracks estimated from X-ray CT scan images

indicated that it took about four times longer for specimens with epoxy coated steel to

yield crack volumes similar to those of specimens with black steel.

• The damage growth rate between specimens with epoxy coated steel and black steel

varied from 2.5 to 4.0 depending on whether the corrosion rate or volume of cracks

was taken as the damage measure. In order to be conservative, it is recommended that

the estimated life expectancy for decks reinforced with epoxy coated rebar is 2.5

times that of decks reinforced with black rebar.

6.4 Final Remarks

The data in this research suggests that the use of ECR in concrete bridge decks will

significantly reduce corrosion induced damage, thereby extending the lifetime of concrete bridge

decks reinforced with black steel. However, there has been much debate regarding this topic. As

mentioned in Section 1.3.2, much previous research agrees with the findings of this thesis that

ECR improves the performance of concrete bridge decks. On the other hand, some research has

concluded that ECR does not improve the performance of concrete bridge decks and may even

negatively impact performance. Therefore, it is important to view the results of this thesis in the

light of a few critical observations on how this research can be applied to concrete bridge decks

in the field. First, the small-scale specimens used are not a realistic representation of a concrete

bridge deck. The simplified layout of the steel reinforcement did not allow for spalling or

delamination of the concrete, which are common in concrete bridge decks. Also, the concrete

used in bridge decks in the field is different than the concrete used in this research. The many

admixtures used in concrete bridge decks in the field will certainly have an impact on the deck’s

71

ability to resist corrosion induced damage. Furthermore, the amount of damage to the epoxy

coating on the steel significantly affects its ability to resist corrosion. A damage amount of 2

percent was used as suggested by the Research Advisory Panel at MDOT. However, this may not

be representative of concrete bridge decks in the field. Although environmental and loading

conditions were the same for ECR and black steel specimens, the laboratory data reported herein

should be validated using field observations.

72

APPENDICES

73

APPENDIX A

STIFFNESS DATA

Table A.1: Black Level 0 Table A.2: Epoxy Level 0 Unrepaired

Cycles Stiffness (k/in)

1 2 3

0 37.8 32.9 28.5

160,000 37.0 31.5 23.3

170,000 41.1 35.7 27.2

245,415 52.6 32.6 46.6

565,279 52.2 33.4 48.4

743,822 53.3 35.1 50.6

1,088,582 52.2 36.5 56.4

1,330,662 57.7 33.0 54.6

1,582,337 57.7 37.8 55.0

1,921,590 52.8 36.3 51.7


1 2

0 60.9 67.1

146,935 56.3 59.8

321,980 51.9 58.2

482,885 68.8 66.1

606,165 76.3 66.2

858,481 64.9 66.8

1,038,572 71.4 56.2

1,193,625 65.0 67.1

1,493,625 68.0 65.6

1,577,485 70.9 66.2

1,756,110 69.6 66.8

1,908,408 66.6 75.2

Table A.3: Black Level 1 Unrepaired Table A.4: Epoxy Level 1 Unrepaired

Cycles Stiffness

Cycles Stiffness

(k/in) (k/in)

1 2

0 57.6 0 57.0

146,935 45.3 320,504 57.3

321,980 53.6 644,222 Failed

482,885 46.2 - -

606,165 51.9 - -

858,481 35.2 - -

1,038,572 61.0 - -

1,193,625 62.3 - -

1,493,625 64.5 - -

1,577,485 66.0 - -

1,756,110 63.7 - -

1,908,408 67.7 - -


1 2 3

0 69.7 62.8 55.4

249,109 75.2 70.7 52.3

588,014 68.7 65.9 50.9

842,443 65.3 72.4 52.8

1,011,581 71.7 89.0 55.5

1,353,276 68.9 82.7 55.7

1,437,992 70.9 72.5 56.5

1,938,583 69.7 80.0 56.9

74

Table A.5: Black Level 1 Repaired Table A.6: Epoxy Level 1 Repaired


1 2

0 67.7 67.4

326,420 72.7 69.6

578,156 72.6 84.2

893,156 73.7 40.0

1,148,889 67.7 69.9

1,483,987 64.3 72.1

1,816,948 60.5 71.7

1,996,324 56.7 73.9


1 2 3

0 79.2 49.3 74.0

236,057 78.9 46.3 70.8

486,801 75.9 46.8 70.6

826,156 70.2 47.9 62.0

1,078,285 77.2 45.7 68.4

1,421,719 75.4 46.9 72.9

1,671,955 75.0 47.4 68.8

1,703,598 72.4 45.1 71.2

2,068,395 75.5 45.7 79.7



1 2

0 58.2 63.6

320,504 76.7 41.7

644,222 71.9 47.9

744,710 Failed Failed


1 2 3

0 72.8 78.5 74.9

243,446 71.3 77.7 79.5

581,414 69.4 83.0 80.1

1178,510 75.7 71.8 78.0

1178,510 74.3 77.4 78.1

1464,193 74.0 80.6 71.0

1,855,593 70.9 86.1 69.3

2,024,904 66.2 87.5 76.5



1

0 53.4

326,420 53.4

578,156 -

893,156 -

1,148,889 54.4

1,483,987 49.9

1,816,948 54.9

1,996,324 48.2


1 2 3

0 66.9 72.3 67.6

166,188 64.1 70.7 48.6

595,398 - - -

836,307 65.2 59.8 45.8

1,156,889 67.3 59.5 43.7

1,401,185 63.6 43.4 Failed

1,436,412 69.9 Failed -

1,779,847 66.3 - -

2,130,928 65.4 - -

75



1 2 3

0 Failed Failed Failed


1 2 3

0 95.2 Failed 69.5

414,523 Failed - Failed



1 2 3

0 - Failed -

239,061 67.1 - 55.1

751,111 62.2 - 53.0

1,053,614 60.8 - Failed

1,739,217 49.5 - -

1,995,881 52.6 - -


1 2

0 57.6 55.9

434,020 68.3 57.7

681,744 63.8 64.1

1,024,890 66.3 57.1

1,279,421 66.2 52.8

1,591,507 56.1 Failed

2,025,925 61.1 -



1 2 3

0 Failed Failed Failed


1 2 3

0 56.2 Failed 74.3

184,708 52.5 - 70.5

596,692 55.8 - 70.7

1,182,266 50.8 - 62.6

1,430,488 51.7 - 56.8

1,602,313 38.9 - Failed

1,676,880 Failed - -

76



1 2 3

0 83.5 96.9 95.1

257,041 109.8 61.4 93.7

583,949 99.1 65.8 91.4

854,187 106.1 65.8 96.5

1,181,715 106.9 65.3 86.6

1,433,461 98.7 64.2 102.9

1,775,923 101.9 56.9 92.3

2,048,680 94.9 63.7 79.2


1 2 3

0 Failed Failed 78.7

332,288 - - 73.4

675,110 - - 75.5

1,275,384 - - 75.7

1,527,247 - - 74.3

1,872,321 - - 60.4

2,123,588 - - 74.5

77

APPENDIX B

X-RAY IMAGES AND PROCESSED CRACK IMAGES

Figure B.1: B-0 X-Ray Image and Crack after Applying Threshold

Figure B.2: E-0 X-Ray Image with No Crack

78

Figure B.3: B-1-N X-Ray Image and Crack after Applying Threshold

Figure B.4: E-1-N X-Ray Image with No Crack

79

Figure B.5: B-1-R X-Ray Image and Crack after Applying Threshold

Figure B.6: E-1-R X-Ray Image with No Crack

80

Figure B.7: E-2-N X-Ray Image and Crack with No Crack


81

Figure B.9: E-2-R X-Ray Image and Crack after Applying Threshold

Figure B.10: B-3-N X-Ray Image and Crack after Applying Threshold

82

Figure B.11: E-3-N X-Ray Image and Crack after Applying Threshold


83



84


85

APPENDIX C

CORROSION CHARGE DATA

Table C.1: B-0 Corrosion Data

Time

(days)

Charge

(Ah)

0 0

2.84 0.24

6.54 0.24

8.61 1.03

12.60 2.34

15.40 3.51

18.31 4.58

22.24 5.71

Table C.2: E-0 Corrosion Data

Time

(days)

Charge

(Ah)

0 0

1.70 0.06

3.73 0.09

5.59 0.14

7.02 0.15

9.94 0.15

12.02 0.30

13.82 0.31

17.29 0.32

18.26 0.32

20.33 0.32

22.09 0.32

Table C.3: B-1-N Corrosion Data

Time

(days)

Charge

(Ah)

0 0

1.00 0.09

2.00 0.46

5.00 2.59

8.00 5.81

11.00 10.67

12.00 12.38

13.70 12.44

15.73 12.47

17.59 12.52

19.02 12.53

21.94 12.53

24.02 12.68

25.82 12.69

29.29 12.70

30.26 12.70

32.33 12.70

34.09 12.70

86

Table C.4: E-1-N Corrosion Data

Time

(days)

Charge

(Ah)

0 0

6.00 0.67

8.00 0.98

11.00 1.59

21.00 2.97

28.00 5.06

31.00 6.33

35.00 8.30

38.00 10.21

44.00 14.08

Table C.5: B-1-R Corrosion Data

Time

(days)

Charge

(Ah)

0 0

1.00 0.08

2.00 0.45

5.00 3.13

800 7.11

11.00 12.72

12.00 15.03

15.78 16.82

18.69 18.04

22.34 18.89

25.30 20.29

29.18 21.93

33.03 23.47

35.11 29.12

Table C.6: E-1-R Corrosion Data

Time

(days)

Charge

(Ah)

0 0

10.00 0.57

12.00 1.13

26.00 3.78

33.00 7.81

39.00 11.17

41.00 12.11

44.00 13.68


Time

(days)

Charge

(Ah)

0 0

1.00 0.08

2.00 0.39

5.00 2.46

8.00 5.38

11.00 9.94

12.00 11.80

13.00 15.59

15.00 18.97

16.00 20.80

20.00 27.60

22.00 31.20

26.00 38.15

29.00 44.82

31.00 50.15

34.71 51.66

38.46 53.12

39.62 54.77

87


Time

(days)

Charge

(Ah)

0 0

1.00 0.01

2.00 0.12

5.00 0.69

8.00 1.36

11.00 2.40

12.00 2.84

13.00 3.32

15.00 4.72

16.00 5.68

20.00 9.38

22.00 11.12

26.00 14.87

29.00 17.95

31.00 19.82

36.00 20.26

40.00 21.06

47.00 23.01

50.00 24.10

55.00 25.94

57.00 26.85

62.00 29.18

65.00 30.66

71.00 33.52

77.00 36.19

84.00 39.56

87.00 40.38

91.00 40.67

101.00 42.09

103.00 43.15

117.00 46.93

124.00 51.45


Time

(days)

Charge

(Ah)

0 0

1.00 0.08

2.00 0.39

5.00 2.46

8.00 5.38

11.00 9.94

12.00 11.80

13.00 13.46

15.00 17.12

16.00 19.25

20.00 26.81

22.00 30.95

26.00 38.63

29.00 46.48

31.00 52.15

34.78 53.94

37.69 55.16

41.34 56.01

44.30 57.41

48.18 59.05

52.03 60.59

54.11 66.24

88


Time

(days)

Charge

(Ah)

0 0

5.00 0.07

9.00 0.41

16.00 1.89

19.00 2.74

24.00 4.25

26.00 5.11

31.00 7.40

34.00 8.93

40.00 12.25

46.00 15.73

53.00 20.27

56.00 21.33

60.00 21.74

70.00 23.15

72.00 24.27

86.00 27.63

93.00 31.87

99.00 35.16

101.00 36.06

104.00 37.63

114.00 40.99

121.00 46.64

121.04 46.66

122.92 46.73

127.89 46.80

130.68 46.89

134.39 46.97

137.22 47.07

137.63 47.13

141.60 47.30

145.66 47.46


Time

(days)

Charge

(Ah)

0 0

3.00 0.60

7.00 2.54

10.00 5.82

16.00 14.09

23.00 28.40

30.00 48.66

36.00 66.51

39.00 76.33

89


Time

(days)

Charge

(Ah)

0 0

1.00 0.01

2.00 0.09

5.00 0.63

8.00 1.30

11.00 2.36

12.00 2.89

13.00 3.43

15.00 4.92

16.00 5.89

20.00 9.75

22.00 11.50

26.00 15.09

29.00 18.10

31.00 19.98

36.00 20.62

40.00 21.98

47.00 24.34

50.00 25.65

55.00 27.60

57.00 28.83

62.00 32.10

65.00 34.07

71.00 38.04

77.00 42.10

84.00 46.79

87.00 47.85

91.00 48.29

101.00 50.27

103.00 51.71

117.00 56.70

124.00 62.69

130.00 68.17

132.00 69.38

135.00 71.50

Cont’d →

145.00 71.50

152.00 77.12

156.80 79.38


Time

(days)

Charge

(Ah)

0 0

10.00 1.36

17.00 6.11

20.00 10.11

24.00 16.22

27.00 21.93

33.00 33.52

40.00 47.72

47.00 63.47

53.00 75.37

56.00 75.38

58.77 79.64

64.69 83.11

68.19 85.72

76.13 89.07

79.10 91.64

90


Time

(days)

Charge

(Ah)

0 0

5.00 0.25

9.00 1.36

16.00 4.44

19.00 6.16

24.00 9.12

26.00 10.69

31.00 14.74

34.00 17.48

40.00 23.25

46.00 28.77

53.00 35.64

56.00 37.22

60.00 37.88

70.00 40.41

72.00 42.27

86.00 48.10

93.00 54.90

99.00 60.50

101.00 61.83

104.00 64.09

114.00 68.86

121.00 75.96

126.02 76.76

128.89 77.42

132.86 78.71

135.81 79.74

139.42 81.16

144.45 82.93


Time

(days)

Charge

(Ah)

0 0

3.00 3.50

4.00 4.70

7.00 8.60

10.00 12.57

14.00 17.91

17.00 22.26

21.00 29.74

24.00 35.52

28.00 44.09

32.00 52.91

35.00 59.56

40.00 67.53

43.00 71.94

46.00 79.31

50.00 89.17

52.00 94.28

56.00 104.74

58.00 109.99

63.00 122.61

91


Time

(days)

Charge

(Ah)

0 0

3.00 0.35

4.00 0.53

7.00 1.34

10.00 2.53

14.00 3.75

17.00 5.09

21.00 6.74

24.00 7.66

28.00 9.02

32.00 10.88

35.00 12.41

40.00 13.58

43.00 14.03

46.00 15.13

50.00 16.87

52.00 17.51

56.00 18.49

58.00 18.88

63.00 20.65

78.00 20.97

80.00 22.74

81.00 24.78

85.00 34.07

87.00 39.57

91.00 49.94

94.00 59.70

96.00 66.12

101.00 66.57

105.00 68.36

112.00 71.73

115.00 73.37

120.00 76.28

122.00 77.93

127.00 82.23

Cont’d →

130.00 85.24

136.00 91.73

142.00 98.68

149.00 107.38

152.00 109.29

156.00 109.92


Time

(days)

Charge

(Ah)

0 0

3.00 4.53

4.00 6.07

7.00 10.81

10.00 15.36

14.00 21.42

17.00 25.91

21.00 32.30

24.00 37.17

28.00 44.22

32.00 51.80

35.00 57.98

40.00 63.74

43.00 66.87

46.00 72.01

50.00 78.74

52.00 82.15

56.00 89.25

58.00 92.93

63.00 102.49

92


Time

(days)

Charge

(Ah)

0 0

1.00 0.01

2.00 0.14

5.00 0.65

8.00 1.33

11.00 2.37

12.00 3.03

13.00 3.74

15.00 5.71

16.00 7.05

20.00 12.05

22.00 14.39

26.00 18.84

29.00 22.71

31.00 25.39

36.00 26.19

40.00 27.82

47.00 31.69

50.00 33.73

55.00 37.89

57.00 39.73

62.00 44.26

65.00 47.23

71.00 53.37

77.00 58.34

84.00 64.42

87.00 65.77

91.00 66.23

101.00 68.35

103.00 69.89

117.00 74.92

124.00 81.41

130.00 85.17

132.00 86.20

Cont’d →

135.00 88.15

145.00 92.37

152.00 98.24

155.00 102.05

159.00 107.46

93

REFERENCES

94

REFERENCES

[1] Samples, L.M. and Ramirez, J.A., Methods of Corrosion Protection and Durability of Concrete Bridge Decks Reinforced with Epoxy-Coated Bars – Phase I. 1999, Purdue University: West Lafayette, IN.

[2] Koch, G.H., et al., Corrosion Cost and Preventive Strategies in the United States. 2002, Turner-Fairbank Highway Research Center: McLean, VA.

[3] Clifton, J.R., Beeghly, H.F., and Mathey, R.G., Nonmetallic Coatings for Concrete

Reinforcing Bars. 1974, Federal Highway Administration: Washington, DC.

[4] Liu, Y. and Weyers, R.E., Modeling the Time-to-Corrosion Cracking in Chloride Contaminated Reinforced Concrete Structures. ACI Materials Journal, 1998. 95(6): pp. 675-681.

[5] Hartt, W.H., et al., Critical Literature Review of High-Performance Corrosion Reinforcements in Concrete Bridge Applications. 2004, FHWA: McLean, VA.

[6] Fraczek, J., Review of Electrochemical Principles as Applied to Corrosion of Steel in a Concrete or Grout Environment. Special Publication, 1987. 102: pp. 13-24.

[7] http://en.wikipedia.org/wiki/Stainless_steel; Accessed on Mar. 7, 2010.

[8] Srensen, B., Jensen, P.B., and Maahn, E., The Corrosion Properties of Stainless Steel

Reinforcement. Corrosion of Reinforcement in Concrete, 1990: pp. 601-610.

[9] McDonald, D.B., et al., Stainless Steel Reinforcing as Corrosion Protection. Concrete International, 1995. 17(5): pp. 65-70.

[10] Weyers, R.E. and Cady, P.D. Cathodic Protection of Concrete Bridge Decks. 1984: ACI.

[11] Brown, M.C., Corrosion Protection Service Life of Epoxy Coated Renforcing Steel in

Virginia Bridge Decks, Ph.D Dissertation, Virginia Polytechnic Institute and State University, 2002

[12] McDonald, D.B., D.W.Pfeifer, and Sherman, M.R., Corrosion Evaluation of Epoxy-

Coated, Metallic-Clad and Solid Metallic Reinforcing Bars in Concrete. 1998, Federal Highway Administration: McLean, VA.

[13] Cui, F. and Krauss, P.D., Corrosion Resistance of Alternative Reinforcing Bars: An accelerated Test. 2006, CRSI: Schaumburg, IL.

[14] Burke, D.F. Performance of Epoxy-coated Rebar, Galvanized Rebar, and Plain Rebar with Calcium Nitrite in a Marine Environment. in International Conference on Corrosion and Corrosion Protection of Steel in Concrete. 1994. England.

95

[15] Swamy, R.N., et al. Durability of Steel Reinforcement in Marine Environment. in 2nd International Conference. 1988. St. Andrews by-the-Sea, Canada.: ACI.

[16] Evaluation of Bridge Decks using Epoxy Coated Reinforcement. 1994, West Virginia Department of Transportation.

[17] Gillis, H.J. and Hagen, M.G., Field Examination of Epoxy-Coated Rebars in Concrete

Bridge Decks. 1994, Minnesota Department of Transportation: St. Paul, MN.

[18] Weyers, R.E. and Cady, P.D., Deterioration of Concrete Bridge Decks from Corrosion of

Reinforcing Steel. Concrete International, 1987: pp. 15-20.

[19] Sohanghpurwala, A.A. and Scannell, W.T., Verification of Effectiveness of Epoxy-Coated

Bars. 1998, Pennsylvania Department of Transportation: Ashburn, VA.

[20] Fanous, F., Wu, H., and Pape, J., Impact of Deck Cracking on Durability. 2000, Center for Transportation Research and Education, Iowa State University: Ames, IA.

[21] Lee, S.K. and Krauss, P.D., Long-Term Performance of Epoxy-Coated Reinforcing Steel in Heavy Salt-Contaminated Concrete. 2004, Federal Highway Administration: McLean, VA.

[22] Smith, J.L. and Virmani, Y.P., Performance of Epoxy-Coated Rebars in Bridge Decks. 1996, Federal Highway Administration: McLean, VA.

[23] Lee, S.K. and Krauss, P., Service Life Extension of Northern Bridge Decks Containing

Epoxy-Coated Reinforcing Bars. 2003, Concrete Reinforcing Steel Institute: Schaumburg, IL.

[24] Brown, D., Epoxy-Coated Reinforcing Steel in Parking Garages. Concrete Reinforcing Steel Institute (CRSI), 1997.

[25] Sagues A.A., Perez-Duran H.M., and Powers R.G. Corrosion Performance of Epoxy-Coated Reinforcing Steel in Marine Substructure Service. in The NACE Annual Conference and Corrosion Show. 1991. Cincinnati, OH.

[26] Sohanghpurwala, A.A., Scannell, W.T., and Hartt, W.H., Repair and Rehabilitation of

Bridge Components Containing Epoxy-Coated Reinforcement. 2002, Transportation Research Board: Washington, DC.

[27] Sohanghpurwala, A.A., Manual on Service Life of Corrosion-Damaged Reinforced

Concrete Bridge Superstructure Elements. 2006, American Association of State Highway and Transportation Officials and Federal Highway Administration: Washington, DC.

[28] ASTM-C666, Standard Test Method for Resistance of Concrete to Rapid Freezing and Thawing. 2003, West Conshohocker, PA: ASTM Internatinal.

96

[29] Harichandran, R. S., and Baiyasi, M. I. "Repair of corrosion-damaged columns using FRP wraps." Report No. RC-1386, Michigan Department of Transportation, Lansing, Michigan. 2000.

[30] MDOT, Special Provision for Thin Epoxy Polymer Bridge Deck Overlay. 2005, Michigan Department of Transportation: Lansing, MI.

[31] MDOT, 2003 Standard Specifications for Construction. 2003, Michigan Department of Transportation: Lansing, MI.

[32] MDOT, Road Design Manual. 2002, Michigan Department of Transportation: Lansing, MI.

[33] AASHTO (2003). LRFD Bridge Design Specifications. American Association of State Highway Transportation Officials, Washington, D.C.

[34] Halleck, Phillip M., Ryan, Timothy, and Stecko, Tim. Penn State University. Center for Quantitative X-Ray Imaging. 2010.

[35] http://www.cqi.psu.edu/about.html; Accessed on June 14, 2010

[36] Kutay, M. E. X-ray CT Image 3D Analysis Tool (XCAT). Michigan State University. 2010.

[37] Rasband, W.S., ImageJ, U. S. National Institutes of Health, Bethesda, Maryland, USA, http://rsb.info.nih.gov/ij/, 1997-2009.

[38] Ang, A.H-S., and Tang, W.H. Probability Concepts in Engineering Planning and Design:

Volume 1, Wiley, New York, 1975.

damage of concrete with epoxy coated reinforcement …

Documents